A nurse manager comes back from a three-day conference with a genuinely good idea. A hospital in another country cut its discharge delays by moving the pharmacy review earlier in the day. She has the slides. She has the presenter’s email. On Monday she mentions it at the team meeting, and everyone agrees it is interesting.
Six months later nothing has changed. Not because anyone opposed it. The pharmacy rota is set by a different directorate. The electronic record would need a new field. The consultant who would have to agree was on leave that week. The idea didn’t fail. It never reached anyone who could make it happen.
This is the normal fate of conference learning, and it is usually misdiagnosed. The standard complaint is that people forget what they heard, so the standard fixes are better slides, recordings and summary notes. But forgetting is not the main problem. The main problem is that conferences transfer knowledge to individuals, while practice and policy are changed by organisations. The attendee comes home knowing something their system has no way to absorb.
Here is the thesis: the knowledge-transfer problem after health events is not a learning problem. It is a delivery problem. Conferences are designed as if the attendee were the destination, when the attendee is only the courier, and nobody has arranged a place to deliver the package.
What implementation science already knows
The field that studies how evidence gets into practice has been saying versions of this for two decades, and its findings are worth taking seriously because they cut against conference habits.
The first lesson is that changing professional practice is a behaviour change problem with several independent parts. Susan Michie, Maartje van Stralen and Robert West’s “behaviour change wheel,” published in Implementation Science in 2011, organises it around three conditions: capability, opportunity and motivation. A conference can raise capability, by teaching people what works and how. It can sometimes raise motivation. It does almost nothing for opportunity: the time, resources, authority and physical and social environment needed to act. For most practice changes in hospitals and health systems, opportunity is the binding constraint.
The second lesson is that context decides whether an intervention takes. The Consolidated Framework for Implementation Research, developed by Laura Damschroder and colleagues and published in Implementation Science in 2009, lists the conditions that shape implementation, including the inner setting of the organisation, its culture, leadership engagement, available resources and readiness for change. The framework has been widely used precisely because implementers kept finding that the same intervention succeeded in one site and stalled in the next. A conference presentation transmits the intervention. It rarely transmits the context that made it work, and it never transmits anything about the context it is going home to.
The third lesson concerns policy. Kathryn Oliver and colleagues published a systematic review in BMC Health Services Research in 2014 on what helps and hinders policymakers’ use of evidence. The facilitators most often cited were relationships, contact and collaboration with researchers, along with access to relevant, reliable and clear research. The barriers were largely the reverse: research that wasn’t relevant or clear, and poor access to it. Notice what is on that list and what isn’t. Relationships and relevance matter. Volume of information does not appear as a facilitator at all.
Insight one: the attendee is rarely the decision-maker
Look at who goes to health conferences and who decides whether practice changes. They overlap less than you would think.
Clinicians attend, but they don’t set rotas, budgets or IT configurations. Programme officers attend, but they don’t sign ministerial circulars. Middle managers attend, but they need approval from people above them who did not hear the talk. The senior decision-makers who do attend often spend their time in bilateral meetings and plenaries, not in the practical sessions where implementable ideas are presented.
So the typical transfer chain has at least one extra link: the attendee must persuade someone who wasn’t there. That is a much harder task than understanding the idea in the first place, and conferences give attendees almost no help with it. A slide deck built to impress a room of peers is a poor tool for convincing a finance director.
Insight two: ideas travel through people, not documents
Among the more robust findings in implementation research is that respected colleagues influence practice. A 2019 Cochrane review led by Gerd Flodgren looked at the effect of local opinion leaders, clinicians recognised by their peers as influential, on professional practice. Across 24 studies, involving more than 3,000 health professionals, opinion leaders alone or combined with other interventions probably improved compliance with evidence-based practice, with a median absolute improvement of 10.8%. The range was wide, from 3.5% to 14.6% across the middle half of studies, but the direction was consistent.
The implication for conferences is direct. The person who comes home from an event is most likely to change practice if they are, or can recruit, someone colleagues already listen to. Sending the most junior person, or the person whose turn it is, maximises personal development and minimises organisational change. Few institutions make that trade-off consciously.
It also explains the Oliver review’s finding about policy. Policymakers use evidence from people they know and trust. The value of a conference for policy change lies less in the plenary than in the relationships that form around it, and those relationships only matter if they are maintained after everyone flies home.
Insight three: the right unit of attendance is a team
If practice change needs capability, opportunity and motivation spread across several people, the logical unit of attendance is not an individual. It is a small group that between them holds the relevant authority.
Consider the discharge example. If the nurse manager had attended with the pharmacy lead and the consultant, the idea would have come home with its three necessary approvers already persuaded. The conversation on Monday would have been about how, not whether. Many quality improvement collaboratives work on exactly this principle, bringing teams rather than individuals together over months, and the logic is sound even where the evidence on specific collaborative models is mixed.
Team attendance costs more per event. It may well cost less per change achieved, which is the number that matters, and which almost no organisation measures.
Insight four: the event should end with a commitment, not a summary
Conferences end with closing plenaries, rapporteur summaries and declarations. These are documents about what was said. What they rarely produce is a record of what anyone intends to do.
A small change in design helps. Ask attendees, before leaving, to write down one specific change they will attempt, who they need to persuade and by when. Better still, ask them to share it with a colleague who will follow up. This does not guarantee action, but it converts a general intention into a concrete plan with a named obstacle, which is the first thing implementation frameworks would ask for.
The uncomfortable incentive
Why do events keep designing for individual learning when the evidence points toward organisational change? Because individual learning is what can be sold, counted and reported.
Registrations are sold one person at a time. Attendance is counted per head. Satisfaction surveys ask individuals whether they found the sessions useful, and they almost always say yes, because the sessions often were interesting. None of these measures asks whether anything changed afterwards, and no event organiser is paid according to what changes afterwards. Employers, for their part, often treat conference attendance as a staff benefit rather than an investment, so they don’t ask for a return either.
Everyone is behaving rationally. The result is an industry that is very good at transmitting knowledge and has no feedback loop at all on whether it was used.
The strongest counterargument
The most serious objection is that this asks too much of a conference. Events are one input among many. They inspire people, expose them to ideas they wouldn’t otherwise meet and build networks whose payoff may come years later, in ways no follow-up survey would capture. A clinician who hears about a new approach today may champion it five years from now when the circumstances are right. Measuring conferences by immediate practice change would miss most of their value and push organisers toward narrow, instrumental programmes.
That is partly right. Long, diffuse effects are real, and not every session should be judged by what changes on Monday. But the argument is often used to excuse the absence of any attempt to support near-term change, which is where the easiest gains are. Both can be true: a conference can inspire over the long run and still help attendees act in the short run. At the moment most do the first and ignore the second.
A worked example (hypothetical)
Imagine a regional health authority that sends staff to four or five external events a year. It decides to change how.
Instead of approving individual requests, it picks two problems it wants to solve: say, late antenatal booking and high rates of missed outpatient appointments. For each event, it asks whether sessions address either problem. Where they do, it sends a small team: someone who runs the service, someone who controls the relevant budget or system and someone clinically respected. Each team agrees in advance which sessions to attend and which presenters to meet.
On return, each team has a fortnight to present one proposal to the authority’s management meeting, with a named owner, a cost and a measure. The management meeting either approves a trial, rejects it with a reason or asks for more information. After six months, the authority reviews which proposals were implemented and whether the measure moved.
Nothing in this example is expensive or novel. The authority might send fewer people to fewer events. But every trip now ends at a decision point, and the organisation learns which events and sessions actually produced changes worth keeping.
What each actor can do
Employers and health authorities can tie conference funding to specific problems, send teams instead of individuals where possible and require a short proposal, not a trip report, within a fixed period after return.
Event organisers can design sessions for implementers: include the costs, staffing and context of the case presented, not just the results. They can offer team registration, end with commitment exercises and follow up with attendees months later to ask what changed.
Presenters can include a slide on what it took to make the change in their setting, including what failed first, and share materials designed for persuading colleagues back home, not just for impressing the room.
Attendees can go with a problem, identify before leaving who at home must agree to any change and bring back a proposal instead of a summary.
Policymakers can treat events as places to build the relationships that the evidence says matter most, and maintain them afterwards.
The takeaway
Health events produce slides rather than decisions because they are built to inform individuals, and practice and policy are changed by organisations. The fix is not better content. It is designing the trip, and the event, around the moment when an idea has to be adopted by someone who wasn’t in the room.
When did something you learned at an event actually change how your organisation works, and what made it get through? Tell us in the comments.
